Engineering Randomness (@EERandomness) header
Engineering Randomness (@EERandomness) profile picture

All Things Engineering. Electrical, Mechanical, Software, Firmware, AI, Security and everything in between. Specialize in custom HW/FW/SW for motor control

Joined April 2022
Photo by Engineering Randomness (@EERandomness)Photo by Engineering Randomness (@EERandomness)Photo by Engineering Randomness (@EERandomness)Photo by Engineering Randomness (@EERandomness)Photo by Engineering Randomness (@EERandomness)Photo by Engineering Randomness (@EERandomness)Photo by Engineering Randomness (@EERandomness)Photo by Engineering Randomness (@EERandomness)Photo by Engineering Randomness (@EERandomness)Photo by Engineering Randomness (@EERandomness)
Ummm... of course it can? Is it not completely obvious that this would work?
Video tampering can mislead every frontier model we tested into harmful actions, including a 100% attempt rate on stabbing a baby doll, producing toxic gas, and pouring water into a toaster. We're presenting at CoRL Late Breaking Results next week. 🧵 You Can't See Everything: Perceptual Attacks on Embodied AI
1
3
355
The Democrat "about face" on data center construction has begun. Their plan: 1: finally publish the real data that we already knew 2: Pass "pauses" that don't actually pause anything 3: Quietly end the "pauses" and let them build with precisely the same terms they are meeting now Politicians know they will crush the economy if they don't allow the building to continue, they simply used it as an issue to push to get elected
.@washingtonpost reports a new joint Harvard University and the University of Pennsylvania study finding data centers have "only a small effect on residential electricity prices." Data centers were responsible for "just 4 percent of the utility rate increases over the last five years." **Weak Connection Between Data Centers & Overall Electricity Prices** "Multiple studies by different researchers and organizations have failed to find a strong connection between data centers and the overall trajectory of U.S. electricity prices in recent years." archive.is/20261106122720/ht…
51
Engineering Randomness retweeted
😬
1,863
1,582
149
22,476
4,284,250
Engineering Randomness retweeted
We should not feel so strange about putting things in space that convert sunlight into intelligence the Earth is something that converts sunlight into intelligence too.
54
111
12
1,743
91,961
Engineering Randomness retweeted
Riley Trettel (Energy & Data Center Development at SpaceX) in yesterdays Cochise Data Center meeting. He claims the project will deliver up to: - 40% off residential power bills for Sulphur Springs members, for the life of the project - $53 million a year to the SSVEC Foundation, for education and the community - $190 million a year in extra capital for Sulphur Springs, for projects or back to members - $200 million a year for AEPCO and its members. About $100 million of that for Sulphur Springs - $3.6 billion in new power lines, paid by SpaceX and given to the co-op. The co-op takes on no debt - $30 to $100 million a year in new property tax, and a 20% cut in everyone else's county property taxes - $250 million, one time, in sales tax. He said it could pay off the jail bond and fund the sheriff, fire districts, and the hospital - $50 billion invested in the project - 1,000 permanent jobs and 5,000 construction jobs - No industrial water. Water for staff and maintenance about equal to 20–30 homes a year - No gas turbines and no power plant on the site. The computers buy power from the co-op grid - Dark-sky lighting, pointed at the ground - Cooling equipment more than 1,500 feet from any neighboring property line or public road This project is clearly extremely positive for the area.
The next SpaceX datacenter will be in Cochise Arizona. ~6 GW spread over 9 buildings. > No industrial water use. > No gas turbines and no on-site power generation. > ~3 miles away from the nearest homes
277
453
102
3,380
682,861
Engineering Randomness retweeted
Cornell engineers have developed a semiconductor material with a record-setting bandgap and turned it into working devices, opening the door to electronics that handle much higher voltages while wasting less energy. Read more: news.cornell.edu/stories/202….
13
68
7
442
38,207
Excellent. Time to start the YIMBY movement YES IN MY BACK YARD
SpaceX just confirmed a 6 GW supercompute campus in Arizona. That's almost half the 13 GW they guided for the end of 2027, now on a map. Do you understand how dumb the FUD is now? 747 acres in Cochise County, next to the Apache Generating Station. Hearing is 9 December. 6 GW is the full build-out, across 4 phases. Phase 1 is 1.5 GW. This is not another Brownsville. Different state. Different permit chain. The Texas freeze does not touch it. They designed it for grid power from the plant next door, through SSVEC and AEPCO. Gas is only backup. Not the own-turbine Colossus model. Dry-cooled. About 7 to 10 acre-feet a year at full build-out. The filing puts that at roughly 20 to 30 Arizona homes. No new pumping from the Willcox basin. $SPCX
96
Time to go bargain shopping
BREAKING: Memory stocks are getting crushed for a 2nd straight day $SNDK -16.9% $WDC -15.2% $RMBS -14.8% $STX -13.7% $MU -13.0% $SIMO -11.6% $NTAP -7.5% Investors fear the new Democratic Congress will slow AI data center buildouts.
2
129
Engineering Randomness retweeted
fracking. bitcoin. data centers. this is my career arc the last 15 years. all wildly unpopular infrastructure builds called out for environmental concerns. all created physical abundance to make america prosperous and strong. the arc of time is long. we keep pushing 🫡
59
30
4
671
48,394
Engineering Randomness retweeted
At this important juncture of American history, I'd like to say few words that may impact countless lives: First, I still believe America 🇺🇸 will continue to lead the world in AI and, soon, in superintelligence. As a US citizen, I deeply hope that it does. However, if somehow America decides to pause AI development and undermine its own future, that will not stop the progress of super intelligence. The world is full of millions of brilliant, talented people who embrace the extraordinary empowerment that AI brings. There are also many countries that will gladly welcome the data centers, investments, and technological infrastructure that America turns away. China, with one of the world’s largest pools of AI talent and rapidly expanding energy and computing capacity, will quickly catch up and surpass the United States in the race toward superintelligence if America chooses to pause or slow down. Imagine the consequences if China becomes the first nation to command virtually unlimited digital supergeniuses and millions of highly intelligent robots. It will gain an overwhelming technological, economic, and strategic advantage over America and the rest of the world, reshaping the global balance of power forever. Yet billions of people around the world will also enter an extraordinary age of abundance, enjoying a quality of life equal to or better than what we have in America today. That's a great thing for humanity overall but not so good for America if we were to fall behind. My key point is this, ultimately, intelligence will continue to advance, and humanity will continue to progress and thrive, with or without America or European countries leading the way. Unfortunately thought, those who bring populist politicians to power to regulate, restrict, and slow technological progress will one day (in few years) deeply regret their choices. By then, however, it will be too late. Throughout history, power has changed hands many times because of disastrous decisions made by leaders and populations. Great civilizations have risen and fallen, often through their own shortsightedness. But this time will be fundamentally different. Once superintelligence becomes the foundation of economic, scientific, and technological power, the advantages of being first will compound at unprecedented speed. Those who fall behind will never be able to catch up. We have only a few years left before the outcome becomes irreversible. For some nations, it will be game over: their chance to become economic leaders will vanish, along with much of their global influence and power. For others, it will mark the dawn of an extraordinary new golden age of prosperity and unprecedented power. And inevitably, some will wield that power against others, just as humans have done throughout history whenever intelligence has given them an advantage. Let this be a final warning to those who fail to recognize what is at stake. Do not let fearmongers and populists convince you to embrace their agenda or accept their vision of a dystopian future. Do not allow fear of what might go wrong to blind you to everything that will go right. At the very least, look at history. One of the greatest reasons America became the world’s leading power was its extraordinary commitment to science, technology, innovation, and attracting and empowering brilliant minds. America prospered because it recognized the transformative power of human intelligence and created opportunities for that intelligence to flourish. And now, we stand at the threshold of the greatest explosion of intelligence in the history of life on Earth. There has never been a worse time to be shortsighted, intellectually complacent, or afraid of progress. And there has never been a better time to think from first principles, challenge conventional wisdom, and embrace the extraordinary possibilities ahead. So let this be my final warning: Embrace superintelligence with every ounce of ambition, courage, and determination we possess. It will be humanity’s greatest opportunity to solve virtually every problem we face, including those we do not yet even know exist. For the first time in the history of our species, we are approaching the ability to transcend our biological limitations, overcome diseases, reverse aging, eliminate scarcity, and free future generations from much of the suffering that humanity has endured for thousands of years. Think about what that means. We have spent the entirety of human civilization struggling against the limitations of our own biology and intelligence. Now, we are finally on the verge of breaking through those barriers. To turn away from that possibility, at the very moment it comes within our reach, will be one of the greatest tragedies in human history. We have a choice: help shape the greatest transformation civilization has ever experienced, or watch others build the future while we argue about whether it should exist. And perhaps, ironically, the best advice for this extraordinary moment is a new twist on my favorite motto: Stay foolish enough to believe the impossible is possible. Stay hungry for more intelligence, more discovery, and more abundance. And above all, never stop believing in and building the future. 🇺🇸
45
64
9
527
31,533
I don't think the market has priced in how much NVIDIA Grace is going to grow. This is a hint. nitter.gq/WOLF_Financial/status/…
ARM $ARM SAYS ARM-BASED CHIPS NOW RUN MORE THAN HALF OF SERVERS AT THE TOP HYPERSCALERS Arm says its server share at the biggest cloud providers more than tripled, from 15% in 2024 to more than 50% in 2025, and it expects more than 60% by the end of 2026 as Google Axion, AWS Graviton, Microsoft Cobalt and NVIDIA platforms ramp. Data center royalties doubled from a year ago. Partners also shipped more than 10 billion Arm-based chips in a single quarter for the first time, and Arm-based PCs are now about 20% of global shipments.
2
128
I am starting to seriously question if the horrendous PR by the hyperscalers and labs is actually a strategy. Are they TRYING to make data centers unpopular for another play at regulatory capture?
1
3
123
Engineering Randomness retweeted
The differing perceptions of AI are wild. The country just voted against data centers on 40+ provisions nationwide and hates AI. (Plenty of legitimate gripes.) But most of those people are just unaware that AI is finding a new cancer cure or treatment weekly, going through decades of peer reviewed data and uncovering mountains of errors, that its correcting, throughout every academic field. Replacing billions of man hours of drudgery. Closest thing to a “miracle” I’ve ever witnessed. When I talk to people outside of tech/VC, they’re just unaware of all of it.
464
200
119
2,115
459,004
Can't find the source at the moment, but a few years back a guy showed you could also geolocate a video by comparing it to the tiny frequency fluctuations in power grids, which introduce noise into the video/audio.
You can actually do this with far greater accuracy using just church bells. Can't quite believe it, but I tested it and it works. Planes need a timestamp and unbroken sound. I tested 56 normal YouTube videos: 0 worked. Bells need a few seconds. The bell ringers' database (Dove's Guide) logs every bell tower in England, 5,896 of them, all unique Hz, with the note of its biggest bell. From a video with few seconds of the audio, the bells in a Derbyshire walk picked out the one church. I'm stunned.
2
5
331
Who does a software update on Election Day?
BREAKING: A software update left some voting machines in New York’s Westchester County unable to read ballots on Election Day, officials say. Voting continued under emergency procedures while officials worked to restore the scanners. County Executive Ken Jenkins has promised a bipartisan investigation.
2
5
336
There is a reason AMD didn't report the datacenter GPU and CPU revenue split. All CPUs are in high demand (and will continue to be), but AMD GPUs are not competitive
1
1
385
This is a form of "AI will invent their own language". Turns out the language may be the weights themselves.
We asked GPT-6 Astra to edit Qwen3.5-0.8B weights to improve math. It directly changed only 0.016% of weights, fixing 0.8B's reasoning loops & improving math up to 6x and even instruction following 3x, without gradients or training. We call this Weight-Space Programming: frontier LLMs can now natively read/write neural network weights! 🧵 Blog: rikiyat.github.io/blog/weigh…
Readers added context they thought people might want to know
LLM was given weights explicitly as text in prompt and used code/forward passes (per authors' logs) to ablate Qwen looping direction, a known thinking-mode issue per model card. Primarily fixed repetition in that mode, not general 'natively' improving weights/math. rikiyat.github.io/blog/weight-sp… github.com/RikiyaT/lm_und… huggingface.co/Qwen/Qwen3.5-0… huggingface.co/RikiyaT/Qwen3.…
1
107
Engineering Randomness retweeted
Does Google realize that the Pavlovian response they’re teaching the world is that model releases from them are Incredibly Annoying? They release something, claim incredible levels of SOTA, have given access to their entire inner circle that You Are not Invited to and then just don’t give access for weeks (?) And then they expect you to like it when they launch stuff? Haha sorry for the rant #carbongate
19
4
197
8,791
Again, compare $/GPU/hr to minimum wage. This will become a very important metric to quantify productivity. If a GPU rents for more than a person's wage, it means the market gains more economic value from it. ....And we can build LOTS of gpus nitter.gq/EERandomness/status/20…
Meanwhile B300 What’s the point of this new FUD now?
1
1
154